Methods of Optical Analysis
Photons from the ultraviolet, optical, and infrared portions of the electromagnetic spectrum have been used for over 100 years to investigate the properties of matter. These techniques, hereinafter referred to as “methods of optical analysis”, include but are not limited to Raman spectroscopy, infrared (IR) spectroscopy, atomic absorption spectroscopy, diffuse reflectance spectroscopy, fluorescence spectroscopy, photoluminescence spectroscopy, and elastic scattering spectroscopy.
In Raman spectroscopy, a sample is irradiated with a substantially monochromatic light source. A small percentage of the incident photons absorbed by the sample are instantaneously re-emitted at slightly different wavelengths. These shifts in wavelength, referred to as Stokes or anti-Stokes shifts, result from changes in the rotational and vibrational states of the constituent molecules. The emitted spectra, captured in a backscattered configuration and analyzed with the spectrometer, reveal very specific information about the chemistry and structure of the sample, particularly information related to carbon-carbon bonds.
IR spectroscopy is similar to Raman, but operates at longer wavelengths. The method is sensitive to functional group vibrations especially OH stretch in water, and is good for studying the substituents on organic molecules. Also, the method can use the unique collection of absorption bands to confirm the identity of a pure compound or to detect the presence of specific impurities
Fluorescence spectroscopy is similar to Raman spectroscopy, in that a sample is irradiated with a substantially monochromatic light source and the re-emitted spectra is captured and analyzed by a spectrometer. However, in the case of fluorescence the emitted spectrum is derived from certain electronic transitions in the sample's constituent molecules. This spectrum is broader and more intense than Raman spectra, lacks Raman's fine structure, and occurs over an extended time frame. Fluorescence spectroscopy is used to determine the chemical constituents of a sample.
In diffuse reflectance spectroscopy (DRS), a broadband light source irradiates a mostly opaque sample. Certain wavelengths of light are selectively absorbed by the sample, and some are scattered. A spectrometer configured to capture backscatter analyzes the spectrum of the scattered rays. “Dips” in the spectrum, caused by absorption in the sample, reveal information about the molecular content of the sample.
In DRS, scattered photons captured by the spectrometer may have undergone elastic or inelastic scattering, or both. Elastic scattering spectroscopy (ESS) is similar to DRS, except that the geometry of the optical system is controlled so that only rays that have undergone high-angle elastic scattering are captured by a spectrometer. Mie scattering theory is then used to analyze the spectra. These spectra reveal information about the size of the scatterers, their index of refraction, the average distance between scatterers, and ranges of values on these measures. This technique has been used to analyze industrial materials such as slurries containing liquids and suspended particles. It has also been used to assess biological tissues. In this case, the ESS spectrum reveals the size of intracellular components such as nuclei and mitochondria. Enlargement of these components above normal levels may indicate a disease state such as cancer.
In atomic absorption spectroscopy, a broadband light source and a spectrometer are arranged in an “opposed” configuration (facing each other) to measure gas-phase atoms. Since the samples of interest are usually liquids or solids, the analyte atoms or ions must be vaporized in a flame or graphite furnace in order to be analyzed. The atoms absorb ultraviolet or visible light and make transitions to higher electronic energy levels. The analyte concentration is determined from the amount of absorption as the light passes through the vaporized sample. This technique is capable of detecting very small concentrations of atoms or molecules in a sample.
Fourier transform (FT) techniques may alternatively be used with many of the optical analysis methods described above. FT techniques convert a time domain measurement to a frequency domain measurement, or vice versa. Instruments employing FT basically reveal the same information about a sample as a comparable instrument without FT, but an FT instrument may be optimized for higher resolution, higher speed, higher sensitivity, or other parameters.
The Use of Polarization in Optical Analysis Techniques
The term “polarization” refers to the spatial orientation of the electromagnetic field associated with each photon. Most naturally-occurring and manmade light sources produce photons with random polarization states. Lasers are generally highly polarized. In all of the optical analysis methods described above, non-polarized light may be used. However, with some of these techniques, the use of polarized light may enable benefits such as higher sensitivities, improved signal-to-noise ratios (SNRs) or additional capabilities.
For example, polarized atomic absorption spectroscopy systems such as the Hitachi Z-5000 offer lower detection limits with a smaller, simpler instrument design compared to non-polarized instruments.
Polarized Raman spectroscopy is used, for example, to determine the secondary and tertiary structures of membrane proteins in biological samples. By studying these aligned proteins with polarized Raman spectroscopy, additional data about the orientation of the bond-polarizability tensors with respect to the known polarization direction of the laser is obtained. This information is combined with molecular models to infer details about the structure of the protein.
In the materials science field, optical strain gauges may also employ polarized Raman. Sensors are constructed by embedding carbon nanotubes in a polymer. Polarized Raman analysis is very sensitive to the strain transferred from the matrix to the nanotubes.
External Reflection Spectroscopy (IRRAS) is used to examine thin films on mirror-like substrates such as coatings and adhesives on metal surfaces. Using a grazing angle technique, the beam makes a high-angle reflection of approximately 88° from the sample and is polarized in the plane of incidence (p-polarization). Polarization sensitivity makes IRRAS useful in determining the orientation of molecules in relation to the metal.
In the polarized variant of ESS (PESS), a polarized broadband source is used to irradiate a sample such as biological tissue. Instead of a single optical channel for measuring the backscattered light in ESS, two channels are used in PESS. One channel is polarized with the same spatial orientation as the source channel, while the other channel is cross-polarized. Since the polarization of backscattered photons depends on the number of scattering events the photon has undergone, and their subtended scattering angle, this detection method collects photons mostly from a well-defined region of the tissue and filters out photons scattered from underlying and surrounding tissue. This enables measurements with high spatial specificity and high signal-to-noise ratio. Recent clinical studies have demonstrated the utility of PESS for the analysis of the surface layers of human tissues lining the outside of the body and body canals (epithelia). Carcinomas originate in epithelial layers, so sampling of this layer independent of the sub-epithelial layers enables the detection of atypical tissues at the earliest stages of growth. ESS and PESS investigations are currently being conducted in many parts of the body, including the gastrointestinal tract (oral cavity, esophagus, stomach, intestines), mammary ducts, bladder, urethra, cervix, and skin.
Fiberoptic Sampling Probes
In many applications, it is desirable to measure a sample in situ, rather than removing a sample from its original location for analysis in a laboratory. Examples in the medical field include measurements of human or animal tissue in vivo, either on the surface of the body, subsurface using a percutaneous technique, or deep inside the body using an endoscope. Pharmaceutical and cosmetic applications include measurements of powders, slurries, suspensions, and solids. Environmental applications include field measurements of water in lakes and streams, and gases in smokestack emissions. Industrial applications include process control measurements in locations such as chemical plants, oil refineries, food processing plants, breweries, and fuel depots. Public safety, security, and forensic applications include detection of explosives residue, illegal drugs, and biohazards such as biological warfare agents, toxic chemicals, and microbial contamination.
In a number of these applications, physical access to the sample is limited. For example, in a lake it may be desired to take a measurement at a depth of 2 meters. In the body, a sample may be required deep in the esophagus. In a cosmetics factory, a sample may be required of a slurry inside a vat or flowing in a pipe. For many of these applications, it is not possible to bring the analytical instrument to the sample. Instead, fiberoptic probes provide the optimal means of conveying light from the instrument to the sample, and/or from the sample to the instrument. Fiberoptic probes are efficient conductors of broadband light, are immune to electromagnetic interference, can be very long (up to hundreds of meters in length), and may be constructed to be flexible, with very small cross sections that can fit into tiny spaces.
Polarized Fiberoptic Probes
Implementing polarized detection in fiberoptic probes has historically had its limitations. Following is a discussion of issues pertaining to the use of fiberoptic probes for Polarized ESS (PESS). However, the main points of the discussion are also applicable to the other optical analysis methods discussed above, and are intended to illustrate the general case.
There are several design difficulties in trying to make small-diameter PESS probes suitable for certain applications, especially in the medical field for needle- or endoscopic-delivery in-vivo.
In order for PESS to work properly, broadband polarized light must be delivered to a sample, and two detection channels must conduct broadband light to spectrometers for analysis of the scattering. The two detection channels must have orthogonal polarizations with a high contrast ratio (at least 10:1, and preferably >100:1). Achieving high contrast detection of two polarization modes over broad passbands is the principal challenge. For the PESS application, “broadband” means a passband of about 600 nm. For other applications, “broadband” may mean as little as 20 nm.
Fabricating polarized optical probes may be approached in two ways. FIG. 1 shows the first approach. An analysis instrument 1 is optically coupled 2 to an optical probe 3. The probe contains one delivery channel 4 and two collection channels 5 and 6. The probe is in optical communication 7 with tissue or another type of sample 8. The polarizers 9 are placed between the probe 3 and instrument 1, and polarize the light as it is transmitted. This is the easier approach because the polarizers are inside the analysis instrument instead of being part of the probe, and so there is little constraint on their size or cost. However, with this arrangement the optical channels 4, 5 and 6 must maintain the polarization of the incident light as the light propagates along the length of the probe 3. If the channels are constructed using conventional optical fiber, polarization is lost, invalidating the measurement. While polarization-maintaining fibers exist, they only offer acceptable performance over a maximum passband on the order of 20 nm. They are thus unsuitable for optical analysis methods employing broadband light.
The second approach is to place the polarizers at the sample end of the probe. This is advantageous, as the light is polarized as it exits the delivery fiber on its way to the sample, and the scattered light is polarized as it enters the detection fibers. Since only light of the correct polarization enters each detection channel, loss of polarization as the light propagates along the length of the channel does not affect the measurement. This allows the probe to be constructed, for example, with commercially available broadband fiber (such as silica-clad-silica).
Nevertheless, this approach presents a number of difficulties. First, for the PESS application, as the probe is mainly intended for measurements of epithelial tissues, it is desirable to confine the sensing volume to the first 300 microns of tissue depth. This constrains the optical geometry of the distal tip—the delivery and collection fibers must be separated by no more than a few hundred microns, and their end faces must be in very close proximity to the tissue. To eliminate crosstalk caused by Fresnel reflections (i.e., light leakage from one optical channel to another caused by reflections from optical surfaces), any polarizer placed between the fibers and the tissue must have a thickness that is less than the spacing between the fibers. This puts further constraints on the size and shape of the polarizers, as well as the fibers.
To be commercially viable, a single-use medical probe must also have a low manufacturing cost. This eliminates the use of any fabricated parts that have a high labor content. If the probe is reusable, the manufacturing cost can be higher, but the probe must withstand high temperature sterilization by steam autoclave. This puts additional constraints on materials, adhesives and coatings.
None of the conventional polarizer technologies has the necessary characteristics for this application. Dielectric thin film cube beamsplitters rely on the difference in the reflectance of S and P polarization states with angle. This means that the surface on which the coating is deposited must be mounted at an angle to the optical axis. This is typically achieved by depositing the polarizing coating onto a 45° surface inside a cube beamsplitter. Unpolarized light is split into S and P components at the 45° surface. However, placing a tiny cube at the sample end of a small diameter probe presents several difficulties. First, because the ray bundles exiting the delivery channels diverge, the cube must be significantly larger than the channel diameter, driving the overall probe size up. This makes its use infeasible in certain applications, especially certain parts of the body. Second, the cross-polarized collection channel requires the use of a second cube that is rotated 90° to the aligned cube. This complicates the design and increases the probe size further; probes using dielectric thin film cube beamsplitters are typically in the range of 1″–4″ in diameter. Third, in order to prevent Fresnel reflections from the sample side of the beamsplitter cubes, those optical surfaces must either be tilted or coated with a high efficiency anti-reflection coating, further increasing complexity and cost. The use of cube beamsplitters thus does not lead to a commercially-viable probe manufacturing cost in any volume
Dichroic Sheet Polarizers have also been used by some researchers to construct fiberoptic probes. “Dichroism” is selective absorption of one polarization plane over the other during transmission through a material. Sheet-type dichroic polarizers are generally manufactured using films of organic materials. The film is stretched, aligning molecules into a birefringent geometry, and then dyed. The dye molecules selectively attach themselves to aligned polymer molecules, so that absorption is high in one plane and weak in the other. The stretched film is then bonded to a transparent substrate or sandwiched between a pair of sheets (glass, plastic, fused silica, etc.) to stabilize it and protect it from the environment. The transmitted beam is linearly polarized. Polarizers made of such material are very useful for low-power and visual applications. The main advantages of this type of polarizer are good performance vs. angle of incidence and thin substrate thickness. However, there are several problems with using dichroic polarizers for fiberoptic probes. First, none of the organic compounds used in these polarizers remain stable when exposed to the temperatures required for steam autoclave (>120° C.). Since steam autoclaving is the most common method of sterilizing reusable medical devices, this is a major impediment to commercialization. Second, none of the dichroic polarizers have the desired spectral bandwidth. Some are optimized for the UV, some for the visible, but none for both. This limits the clinical utility of the probe. Third, though the dichroic sheet polarizers have the shortest optical path of any polarizer, the commercially available ones are still too thick (˜200 microns—Polaroid Corp.). This results in Fresnel reflections from the sample side of the polarizer. Fourth, it is difficult to cut, handle, and bond such a tiny disk of filter material to the end of a fiber probe. Because the films have been stretched in one axis, when mounted on thin substrates, these polarizers curl very strongly, contributing to difficulties in handling. This is exacerbated by the fact that tiny pieces of the film must be mounted with their polarization axes orthogonal to each other. Fifth, since the bonding area is so small, the polarizers could detach, especially if prone to curling, and especially if the probe is used multiple times. If they detached in-vivo, this could pose a health hazard.
Birefringent polarizers have also been considered, but the polarizing crystals are typically in the range of several millimeters thick, making them impractical for small diameter probes. They are also too fragile to be placed at the distal end of a probe, and are too expensive for typical commercial applications.
Wire grid array polarizers use a periodic series of parallel wires etched or deposited onto a substrate, as described in U.S. Pat. No. 6,122,103. This array passes one polarization mode and reflects the orthogonal one. It offers very high transmission, high extinction, large acceptance angle, high temperature tolerance, and may be used normal or inclined to the optical axis. Semiconductor lithographic methods are used to form the array of wires on the transparent substrate. However, commercial wire grid polarizers are typically deposited on substrates several mm thick. Using such a thick polarizer at the distal end of a small diameter probe would result in unacceptable performance due to Fresnel reflections. Also, wire grid polarizers have been used mainly in the display industry. Thus, the substrates typically have a large area, all with the same polarization orientation. This arrangement is called a “sheet polarizer”.
In sum, none of the existing polarizer technologies have the combination of bandwidth, small size, thinness, low cost, durability, sterilizability, or pixelability to work in a small diameter probe.